Effective detection and identification of sheath-tailed bats of Australian forests and woodlands

Abstract

Assessing the risk to threatened species of population decline from anthropogenic disturbances is challenging when there are issues with species identification, and little is known of their biology, distribution, population size, and habitat preference. The bare-rumped sheath-tailed bat (Saccolaimus saccolaimus) is one such species that has a poorly defined distribution over two broad areas of northern Australia. Environmental impact assessments are expected to consider the possibility of its presence in intervening areas outside the known distributions. Our study presents new empirical data that can assist with detection of S. saccolaimus across the entire expanse of northern Australia, provides a critical analysis of acoustics-based identification of the species, and assessed presence within the potentially high value habitat of tall Eucalyptus tetrodonta-dominated forest on the western side of Cape York Peninsula using a combination of trapping and acoustic recordings. Capture of other Saccolaimus species was the greatest of any survey conducted to date in Australia, demonstrating that the capture of these high-flying bat species in tall forest habitats can be relatively effective with mist net arrays hoisted into the tree canopy. In addition, reference echolocation call collections from the focal trapping area plus other locations across northern Australia allowed characterisation and comparison of the calls of most low-frequency-emitting (LFE) echolocating bat species of northern Australia. In addition to separation of species-specific search phase call types using multivariate statistics, a compilation of features from search phase, approach phase and feeding buzz echolocation calls will help distinguish S. saccolaimus from most other LFE species. However, the similarity of the echolocation calls of S. mixtus and S. saccolaimus prevented them from being distinguished from one another. A multi-method approach that emulates the present study and incorporates our recommendations and cautions will lead to robustness in ecological studies and greater clarity in environmental impact assessments.

Observations on the utilisation of a restored wildlife corridor by echo‐locating microbats in North Queensland's Wet Tropics

“…The species is possibly known from other tropical upland sites, but there are no records within the peer‐reviewed literature, nor lodged with relevant databases, so this paper represents the first published record. However, few records probably reflect the paucity of local microbat surveys and the difficulty of detecting the species prior to recent developments in echolocation call identification (e.g., Armstrong et al., 2021). A spectrogram of this record is provided (Figure 2).…”

“…A Decision Tree analysis (GF unpubl data) was applied to all remaining files to group, and tentatively label, detected bat calls based on a combination of call metrics derived from zero‐crossing analysis, for example characteristic frequency (Fc), pulse duration (Dur), time between pulses (TBC), slope (S1 and Sc) and pulse curvature. Preliminary call identities applied by the Decision Tree were then confirmed or adjusted manually by comparing the call spectrograms and derived metrics with those of reference calls from northern Queensland (GF unpubl data), and/or with published call descriptions (e.g., Reinhold et al., 2001; Milne, 2002; Armstrong et al., 2021).…”

Bat research in Australasia – in memory of Les Hall, part 2

“…These kinds of long-term studies are much needed to improve our understanding of Australasia's bats. Armstrong et al (2020a) investigated the detection and identification of sheath-tailed bats, focussing on the threatened bare-rumped sheath-tailed bat (Saccolaimus saccolaimus) in northern Australia. Using a combination of trapping and acoustic recordings, they present new empirical data and critical analysis to assist with acoustic detection and identification of S. saccolaimus, and assess the presence of the species within potentially high-value habitat on Cape York Peninsula.…”